Much of what we learn about politics in political science books is either wrong or highly misleading. I had reached this somewhat depressing conclusion long before I completed my previous book, The American Direct Primary, but my research for that book confirmed that my judgement was largely correct. The problem lies at the heart of how political science is conducted, especially in the United States, and is not the result of political scientists being either fools or knaves. In my experience the vast majority of its practitioners fall into neither of these categories. Rather, the problem is caused by the role played by fashion, especially fashions in methodology, which has two adverse consequences for the development of our knowledge about the political world.

First, the obsession with making one's research compatible with the latest fashion in methodology, together with changes in those fashions, means that research is often conducted without any knowledge of similar work that was undertaken years ago. Thus, the equivalent of the wheel is often reinvented by political scientists about every quarter of a century. Secondly, as fashion changes, so do some conceptions and arguments go unchallenged and become part of the “evident truths” of the profession; they are held to be true because previous generations of political scientists believed them to be true. This was what I had discovered about the origins of direct primary elections in America – the events that led to their introduction actually bore no relation to the standard account that was found in every reference to the subject.